Latent Lip Print Identification Using Fast Normalized Cross-Correlation

Abstract

This paper presents a method of latent lip print identification. Lip prints, among other latent traces, can be found at a crime scene. They are captured by police technicians and after further analysis provide a valuable evidence. In the proposed method image standardization and lip print pattern extraction are performed. Then analyzed lip print is identified using image matching based on Fast Normalized Cross-Correlation. Tests were conducted on 300 lip print images acquired from 50 individuals. The results obtained show promising identification performance and suggest further research.